Research Article Cancer Prevention Research Genome-Wide … · sistent deletion of 9p21 across...
Transcript of Research Article Cancer Prevention Research Genome-Wide … · sistent deletion of 9p21 across...
Rese
GenBarA HArra
Jian GWayn
Abst
Intro
EsosixthStatesfromcanceproximadvanare la
AuthorMedica4ThoraM.D. AGastroe
Note:Preven
CorresversityBoulev4657; E
doi: 10
©2010
Cance1176
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
Cancer
eventionesearcharch Article
ome-Wide Catalogue of Chromosomal Aberrations inrett's Esophagus and Esophageal Adenocarcinoma:igh-Density Single Nucleotide Polymorphism
PrR
y Analysis
u1, Jaffer A. Ajani2, Ernest T. Hawk1, Yuanqing Ye1, Jeffrey H. Lee3, Manoop S. Bhutani3,
e L. Hofstetter4, Stephen G. Swisher4, Kenneth K. Wang5, and Xifeng Wu1ractTo
genesiat eachigh-gloss oaverag4.82 ×1.8%,was li9p21was sesmallMGC4in theAmpl
ced-stagergely inef
s' Affiliatiol Oncologycic and Cnderson Cnterology a
Supplemention Researc
ponding Auof Texas M.ard, Housto-mail: jiang
.1158/1940-
American A
r Prev Res
wnloaded
better understand the molecular mechanisms behind esophageal adenocarcinoma (EAC) tumori-s, we used high-density single nucleotide polymorphism arrays to profile chromosomal aberrationsh of the four sequential progression stages, Barrett's metaplasia (BM), low-grade dysplasia (LGD),rade dysplasia (HGD), and EAC, in 101 patients. We observed a significant trend toward increasingf chromosomes with higher progression stage. For BM, LGD, HGD, and EAC, respectively, thee numbers of chromosome arms with loss per sample were 0.30, 3.21, 7.70, and 11.90 (P for trend =10−7), and the mean percentages of single nucleotide polymorphisms with allele loss were 0.1%,6.6%, and 17.2% (P for trend = 2.64 × 10−6). In LGD, loss of 3p14.2 (68.4%) and 16q23.1 (47.4%)mited to narrow regions within the FHIT (3p14.2) and WWOX (16q23.1) genes, whereas loss of(68.4%) occurred in larger regions. A significant increase in the loss of other chromosomal regionsen inHGDand EAC. Loss of 17p (47.6%)was one of themost frequent events in EAC.Many recurrentregions of chromosomal loss disrupted single genes, including FHIT, WWOX, RUNX1, KIF26B,8628, PDE4D, C20orf133, GMDS, DMD, and PARK2, most of which are common fragile site regionshuman genome. However, RUNX1 at 21q22 seemed to be a potential tumor suppressor gene in EAC.ifications were less frequent than losses and mostly occurred in EAC. 8q24 (containing Myc) and1 (containing CTSB) were the two most frequently amplified regions. In addition, a significant trend
8p23.toward increasing amplification was associated with higher progression stage. Cancer Prev Res; 3(9); 1176–86.©2010 AACR.
in thebaseneedits earclinicaMo
geal sqma (E
duction
phageal cancer is the eighth most common and themost lethal cancer in the world (1). In the United, an estimated 16,470 new cases and 14,530 deathsthis disease were expected in 2009 (2). Esophagealr tends to have a very poor prognosis because ap-ately two thirds of patients who are diagnosed have
disease, at which point current therapiesfective (3). The overall 5-year survival rate
less thEAC ithe pa60% oMostmetapepithecolum1% to125-fotients∼0.5%robuscost-e
ns: Departments of 1Epidemiology; 2Gastrointestinal; 3Gastroenterology, Hepatology, and Nutrition; andardiovascular Surgery, The University of Texasancer Center, Houston, Texas; and 5Division ofnd Hepatology, Mayo Clinic, Rochester, Minnesota
tary data for this article are available at Cancerh Online (http://cancerprevres.aacrjournals.org/).
thor: Jian Gu, Department of Epidemiology, The Uni-D. Anderson Cancer Center, Unit 1340, 1155 Presslern, TX 77030. Phone: 713-792-8016; Fax: [email protected].
6207.CAPR-09-0265
ssociation for Cancer Research.
; 3(9) September 2010
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org from
Surveillance, Epidemiology, and End Results data-is 16.8% (3). These dismal statistics highlight theto develop methods to detect esophageal cancer inly stages and to identify biomarkers that can predictl outcomes.re than 90% of esophageal cancers are either esopha-uamous cell carcinomas or esophageal adenocarcino-AC; ref. 3). Once considered a rare tumor (representingan 5% of esophageal cancers in the United States),s the cancer with the fastest increasing incidence inst three decades and currently accounts for more thanf new esophageal cancer cases in this country (4, 5).EAC cases arise from Barrett's esophagus, or Barrett'slasia (BM), a precursor lesion in which the squamouslium of the esophagus is replaced by a metaplasticnar epithelium. BM is estimated to be present in2% of the general population and confers a 30- told increased risk of developing EAC. However, in pa-with BM, the absolute risk of developing EAC is onlyper patient-year, which calls for more accurate and
t prediction of who may develop EAC to increase theffectiveness of surveillance strategies such as routine
. ebruary 1, 2020. © 2010 American Association
endosnant pof stato hignomamay bthis riplasiavariabkers tcompThe
role dPrevioresolution,(SNP)aberraA numand op16, pand Csistenand thstageshighlysize. Asomalnot oEAC bbiompotenWh
powervariatincludphysicinputthe prdetectbackgperforbiops317Kchromigenes
Mate
TissuA to
diseasfromter weat thesies ustitut
pathothe hparaffwere bscribewerecorreslectedthe edexpertto theManuBloodturer's
SNP aSNP
Hap3prototracted37°Cprecipand hnight.DNAusedspeciflyzedgeneratissueeach Ssus itssinglehigheof logtestedto ideThe loabsoluin tumusingdiseasand dwith
SNP Array of Barrett's Esophagus and EAC
www.a
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
copy among high-risk BM patients (6, 7). The malig-rogression of BM follows a generally accepted seriesges, from metaplasia, to low-grade dysplasia (LGD),h-grade dysplasia (HGD), and finally, to adenocarci-. The risk of developing EAC in patients with HGDe higher than 10% per patient-year (7, 8). However,sk has been hard to assess because the grading of dys-is subjective and there is relatively high interobserverility in its diagnosis. Independent objective biomar-hus may improve the assessment of EAC risk bylementing pathologic grading.accumulation of genetic aberrations plays a pivotaluring the malignant progression from BM to EAC.us studies using candidate region analysis, low-tion conventional comparative genomic hybridiza-and low-density single nucleotide polymorphismarrays have identified many of the chromosomaltions involved in the progression of EAC (9–15).ber of well-known tumor suppressor genes (TSG)ncogenes have been implicated in EAC, including53, p21, APC, Rb, SMAD4, Myc, K-ras, EGFR, cyclins,DKs (16–18). However, except for the most con-t deletion of 9p21 across different histologic stagese loss of heterozygosity (LOH) of p53 in the later, the results for other chromosomal aberrations areheterogeneous in terms of stage, frequency, andhigh-resolution genome-wide profiling of chromo-aberrations in different stages from BM to EAC maynly elucidate the mechanisms of tumorigenesis ofut also identify predictors of malignant progression,arkers of prognosis and treatment response, andtial targets for prevention and therapy.ole-genome high-density SNP array analysis is aful new technology for detecting both copy numberions and LOH events (19). Other notable advantagese its high resolution, its capability of profiling bothal and genetic aberrations, its low levels of requiredDNA (as little as 10 ng), and its high sensitivity inesence of normal cell contamination (allowing theion of LOH in paired samples with ∼67% normalround cells; ref. 19). Nancarrow et al. (20) recentlymed a high-density SNP array analysis of 23 primaryies of EAC tumors. In this study, we used Illumina'sSNP array to profile and compare genome-wide
osomal aberrations in the four stages of EAC tumor- filterinusingsampthen aimplechangon thechangas regichromrepresratio.
is, BM, LGD, HGD, and EAC.
rials and Methods
e samples and DNA extractiontal of 101 (20 BM, 19 LGD, 20 HGD, and 42 EAC)e tissues and their paired normal tissues obtainedThe University of Texas M.D. Anderson Cancer Cen-re included in this study. All tissues were snap frozentime of diagnostic or therapeutic endoscopic biop-
sing a tissue collection protocol approved by the in-ional review board. Experienced gastrointestinalannotfurthe
acrjournals.org
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
logists at M.D. Anderson Cancer Center performedistologic readings of the corresponding juxtaposedin-fixed specimens. The diagnoses of LGD and HGDased on each biopsy using the criteria previously de-d (21). All disease tissues in this cross-sectional studyfrom separate patients diagnosed as having EAC. Aponding normal squamous tissue sample was col-from a healthy-appearing mucosa at least 3 cm fromge of the apparent tumor from each patient by angastroenterologist. The tumors were staged accordingAmerican Joint Commission on Cancer Staging
al (22). DNA was extracted using QIAamp DNeasyand Tissue kit (Qiagen) according to the manufac-instructions.
rray analysisarray analysis was done using Illumina's Human-
00 BeadChip array according to the manufacturer'scol (Illumina). Briefly, 200 ng of genomic DNA ex-from the tissues were denatured and amplified at
overnight. The amplified DNA was fragmented anditated at 4°C, resuspended in hybridization buffer,ybridized to HumanHap300 chips at 48°C over-The unhybridized and nonspecifically hybridizedwas then washed away, and the captured DNA wasas template for one-base extension of the locus-ic oligos on the BeadChips. All SNP data were ana-and exported by BeadStudio 2.0 (Illumina), whichted whole-genome profiles for disease and normals based on the log2R ratio and allele frequency ofNP. We used paired analysis (each disease tissue ver-own adjacent normal tissue reference) instead of anormal reference because paired analysis offers
r sensitivity, better quality of data, and lower variation2R ratio in the case of limited input DNA (19). Wethe built-in autoscoring algorithms of BeadStudiontify LOH, homozygous deletion, and amplification.g2R ratio between tumor and normal samples and thete value of the difference between B allele frequencyor and normal samples (|dAllelefreq|) were obtainedthe BeadStudio software. In paired analysis, bothe and normal samples came from the same patients,eviations of |dAllelefreq| from zero indicated regionschromosomal aberrations. We first performed ag process to select the candidate points for analysisB allele frequency (between 0.4 and 0.6) of normalles following the BeadStudio LOH User Guide. Wepplied a circular binary segmentation algorithm (23)mented within the R software environment to identifye points for the regions of aberration using |dAllelefreq|candidate points. Regions that were bound by thesee points and that deviated from zero were identifiedons of chromosomal aberrations. For these regions ofosomal aberrations, we then determined whether itented a chromosomal loss or gain using the log2RThe resulting annotations were compared with the
ations from the built-in BeadStudio algorithms andr confirmed by manually inspecting the BeadStudioCancer Prev Res; 3(9) September 2010 1177
. ebruary 1, 2020. © 2010 American Association
GenomalleleLOHdeleti(a) howere hficulthomoalthouregionthe XfrequchrompercenfromSNPs
StatisFor
separain theeach ctissuewere umosoeach tsize ofwith aKruskcomp
percenand Estatistsided,
Resu
PatienA to
diseaTablewhompatienyears)subjestagesLGD(57.62lattercasianfollowand 3entiat
GenoWe
ing Ill∼317
Tab
Var
AgeSex
MF
EthnCO
StagIIIIIV
StagGra
M
P
U
*P f†P f
Gu et al.
Cance1178
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
e Viewer plots of tumor B allele frequency, normal Bfrequency, |dAllelefreq|, and log2R ratio. We grouped(including copy-neural LOH) and homozygouson together as chromosome loss for two reasons:mozygous deletion was a rare event; (b) when thereigh levels of normal cells in disease tissues, it was dif-to differentiate among LOH, copy-neutral LOH, andzygous deletion because the log2R was ambiguous,gh it was clear from the |dAllelefreq| plot that thehad one of these events. We manually inspected
chromosome from each tissue and identified high-ency aberrations; however, we only counted auto-osomes in the analyses of total number, size, andtage of aberrations. We excluded Y chromosomes
the analysis because there are only two representingon it.
tical analysesall analyses, loss and amplification were analyzedtely. Fisher's exact test was used to test for differencesfrequency of aberrations (loss and amplification) byhromosome arm among BM, LGD, HGD, and EACs. The Kruskal-Wallis and nonparametric trend testssed to compare the sizes of aberration at each chro-me arm among BM, LGD, HGD, and EAC tissues. Forissue sample, we also tallied the total number and totalchromosome aberrations, and the percentage of SNPsberrations among all the analyzed SNPs (317K). The
al-Wallis and nonparametric trend tests were used toare the mean total number, mean total size, and meanHap3(excep
le st and t cteristic
iab
(y 54.60 62.21
alem 8 1icauth 1 1e
I
edeodd
rom Fisher's exact test.
r Prev Res; 3(9) September 2010
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
tage of SNPswith aberrations among BM, LGD,HGD,AC tissues. All analyses were done using the Stata 8.0ical software package (Stata Co.). All tests were two-and P < 0.05 was considered statistically significant.
lts
t and tumor characteristicstal of 101 (20 BM, 19 LGD, 20 HGD, and 42 EAC)se tissue specimens were included in this study.1 lists selected characteristics of patients fromthese tissues were obtained. BM tissues were fromts who were younger (mean age ± SD: 54.60 ± 10.34and included a higher percentage (40%) of femalects than patients with the other three histologic. The mean age was similar among patients with(62.21 ± 9.82), HGD (61.70 ± 12.62), and EAC± 9.51). More than 90% of patients in each of these
three stages were men and more than 80% were Cau-. The distribution of EAC patients by stage was ass: 17 with stage II, 20 with stage III, 2 with stage IV,unspecified; 21 patients had moderately differ-
ed and 20 had poorly differentiated tumors.
me-wide catalogue of chromosomal aberrationsprofiled genome-wide chromosomal aberrations us-umina's HumanHap300 Beadchips, which containedK haplotype tagging SNPs. SNPs on the Human-
00 chips represent even coverage by each chromosomet the Y chromosome) and are evenly spaced across1. Selected ho
umor chara s61.70
2
4
. ebruar
57.62
3
8
17
3
21
Cancer
y 1, 2020. © 2010 Am
0.067*
Prevention Resear
erican Association
les
n (%) n (%) n (%) n (%) BM, LGD, HGD, EAC, P), mean (SD)
(10.34) (9.82) (12.62) (9.51)e 12 (60.0) 18 (94.7) 18 (90.0) 39 (92.9) 0.006†
ale
(40.0) (5.3) (10.0) (7.1) ity casian 19 (95.0) 18 (94.7) 16 (80.0) 34 (81.0) 0.275†er
(5.0) (5.3) (20.0) (19.0)(40.5)
(47.6) 202 (4.8)
Unspecified (7.1)eratelyifferentiated
(50.0)
rly
(47.6) oodifferentiated20
nspecified 1 (2.4)
rom ANOVA test.
ch
the whaveragand thing be
on th9 kb a
Fig. 1.aberrattissue sY axis dp-arm aeach chrepreseambigu1 to 7;15 to 2and amplifications are marked.
SNP Array of Barrett's Esophagus and EAC
www.a
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
ole genome to ensure comprehensive coverage. One, there is one SNP every 9 kb across the genome,
e median spacing is 5 kb. The 90th percentile spac-tween loci, indicating the largest intervals in contentlutionaberra
acrjournals.org
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
e chip, is 19 kb (24). With a mean SNP spacing ofnd a 10 SNP smoothing window, the effective reso-
Genome-wide catalogue of chromosomalions on each chromosome. X axis denotesamples, grouped by histologic stages.enotes chromosomes, oriented with thet the top and q-arm at the bottom forromosome. Red represents loss, greennts amplification, and black representsous aberrations. A, chromosomes (Chr.)B, chromosomes 8 to 14; C, chromosomes2. Some of the most frequent losses
of HumanHap300 SNP array for chromosomaltion analysis is ∼90 kb (19).
Cancer Prev Res; 3(9) September 2010 1179
. ebruary 1, 2020. © 2010 American Association
Figuand asual eual inin histeredand 1quent47.4%in the(Fig. 116q23sivelyand Wthe mregionIn t
loss (creaseLGD9p21of 9pWe
aberraber, toand coAs shoing nucreasiwith iers of
3.21,spectiBM, LThe mwere8.13 ×withfor BM2.64 ×Sim
less fraveragper saHGD,ever, tmalig0.147The m11.0010−8; FficatioHGD,
IndivoverlChr
finedever, tand gmostwere(30%11q,chrom16q ((38.1(38.1%(35.7%osomstagewere 1(P = 5locus)(>20involvp53 anThe
rupte(16qMGC(20p1(6q26man g21q22LGD alost th
Tablearm (
Chrom
3p4p4q5q6p8p9p11p11q12q16q17p18q19p21q22q
NOTinclu*Fis
Gu et al.
Cance1180
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
re 1 shows the genome-wide distribution of lossesmplifications by chromosome in each sample. A vi-xamination of the figure reveals that there is a grad-crease in chromosome aberrations with the increasetologic stages. In BM samples, there were only scat-chromosome losses at 3p14.2 (10%), 9p21 (5%),6q23.1 (5%); these three events were the most fre-aberrations in LGD samples (68.4%, 68.4%, and, respectively; Table 2). There were dramatic increasesnumbers of aberrations in the HGD and EAC stages; Table 2). Furthermore, the 3p14.2 (68.4%) and.1 (47.4%) losses seen in the LGD stage were exclu-limited within genes FHIT (3p14.2; Figs. 2A and 3A)WOX (16q23.1; Supplementary Fig. S1A), two ofost frequently activated common fragile site (CFS)s in the human genome.he HGD and particularly EAC stages, the sizes of 3pFig. 2A) and 16q loss (Supplementary Fig. S1A) in-d. In contrast, 9p loss occurred in larger regions fromto EAC (Fig. 2B), and homozygous deletion of thelocus was observed in the context of large regionLOH (Fig. 3B).then computed the overall levels of chromosomaltion in each sample by three parameters (total num-tal size, and percentage of SNPs with aberrations)mpared the values among the four different stages.wn in Fig. 4, there was a significant trend of increas-mber and size of chromosomal aberrations and in-ng percentage of SNPs with aberrations associated
her's exact test.
ncreasing histologic stages. The average total numb-chromosome arms with losses per sample were 0.30,
in EACis a p
r Prev Res; 3(9) September 2010
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
7.70, and 11.90 for BM, LGD, HGD, and EAC, re-vely (P for trend = 4.82 × 10−7; when analyzingGD, and HGD only, P for trend = 4.44 × 10−5; Fig. 4A).ean total sizes of chromosome losses per sample
2.33, 39.10, 167.87, and 449.90 Mb (P for trend =10−8; Fig. 4B), and the mean percentages of SNPs
allele losses were 0.1%, 1.8%, 6.6%, and 17.2%, LGD, HGD, and EAC, respectively (P for trend =10−6; Fig. 4C).ilar analyses were done for amplifications, which wereequent than losses and mostly occurred in EAC. Thee numbers of chromosome arms with amplificationsmple were 0.30, 0.42, 1.90, and 8.50 for BM, LGD,and EAC, respectively (P for trend = 1.35 × 10−9); how-he trend was not significant for amplification in pre-nant (BM, LGD, and HGD) stages (P for trend =) due to the rarity of amplifications in these stages.ean total sizes of amplifications per sample were, 9.41, 37.70, and 215.74 Mb (P for trend = 1.23 ×ig. 4B), and themean percentages of SNPswith ampli-ns were 0.5%, 0.4%, 1.5%, and 8.0% for BM, LGD,and EAC, respectively (P for trend 1.26 × 10−8; Fig. 4C).
idual chromosome aberrations and smallapping regions of aberrationsomosome losses in LGD samples were mostly con-to three regions (3p14.2, 9p21, and 16q23.1). How-here was a significant increase in chromosome lossesains in the HGD and EAC stages (Fig. 1; Table 2). Thefrequent chromosome arm losses in the HGD stage3p (70%), 9p (50%), 5q (35%), 21q (35%), 6p), 16q, 17p, and 18q (all 25%), and 4q, 8p, 11p,and 12q (all 20%). In EAC, the frequent (>35%)osome losses included 17p (47.6%), 3p (47.6%),42.9%), 4q (42.9%), 6p (40.5%), 9p (38.1%), 8p%), 5q (38.1%), 19p (38.1%), 11p (38.1%), 4p), 11q (35.7%), 18q (35.7%), 21q (35.7%), and18q; Table 2). All of these frequent individual chrom-
e losses showed significant associations with highers; the most significant individual associations7p (P = 2.14 × 10−5), 6p (P = 3.33 × 10−5), and 4p.38 × 10−5; Table 2). The 17p loss (all including p53occurred predominantly in either a large portion of
Mb) or the entire 17p arm, with only one exceptioning a small focal LOH at 17p13.1 (323 kb, containingd 22 other genes; Supplementary Fig. S1B).re were many small recurrent regions of loss that dis-d single genes, including FHIT (3p14.2), WWOX23.1), RUNX1 (21q22.12), KIF26B (1q44),48628 (4q22.1), PDE4D (5q11.2), C20orf1332.1), GMDS (6p25.3), DMD (Xp21.2), and PARK2), the majority of which were CFS regions in the hu-enome (Table 3). Of particular interest was RUNX1 at.12; the focal loss of this gene occurred in two cases ofnd three cases of HGD, and an additional HGD casee entire 21q arm. 21q loss was also a frequent event
2. Frequent loss%)
of ind
ividua l chrom osom eosome arm BM
LGD HGD EAC P*10.0
68.4 70.0 47.6 1.85E− 04 0.0 0.0 10.0 38.1 5.38E− 05 0.0 5.3 20.0 42.9 1.19E− 04 0.0 10.5 35.0 38.1 1.09E− 03 0.0 0.0 30.0 40.5 3.33E− 05 0.0 5.0 20.0 38.1 6.08E− 04 5.0 68.4 50.0 38.1 2.35E− 04 0.0 5.3 20.0 38.1 6.08E− 04 0.0 5.3 20.0 35.7 1.25E− 03 0.0 0.0 20.0 35.7 2.27E− 04 5.0 47.4 25.0 42.9 5.52E− 03 0.0 5.3 25.0 47.6 2.14E− 05 0.0 10.5 25.0 35.7 3.59E− 03 0.0 0.0 15.0 38.1 5.38E− 04 0.0 15.8 35.0 35.7 4.05E−030.0 0.0 10.0 35.7 1.37E−04E: Only losses with >35% frequency in EAC wereded.
(occurring in 35.1%of cases), suggesting thatRUNX1otential TSG involved in the development of EAC.
Cancer Prevention Research
. ebruary 1, 2020. © 2010 American Association
ThecontaiwereThe hcationcandi(36.6%and 1
20q1(25.6and 918q11q13.3ticula
Fig. 2. Dposition
SNP Array of Barrett's Esophagus and EAC
www.a
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
small overlapping regions of amplificationsned from 4 to nearly 100 genes, among which theremany known oncogenes and proliferation genes.igh-frequency small overlapping regions of amplifi-s (occurring in more than 15% of samples) anddate genes in these regions included 8q24.21
istribution of aberrations at chromosomes 3 (A) and 9 (B) in different stages fromin Mb. Red, loss; green, amplification; black, aberrations ambiguous; and dotte
acrjournals.org
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
3.2 (28.6%, ZNF217 and 3 others), 7p11.2%, EGFR and 33 others), 3q26 (20.5%, PIK3CA3 others), 12p12.1 (19.5%, KRAS and 6 others),.2 (17.5%, CTAGE1 and 3 others), and 11q13.2-(16.7%, CCND1 and 10 others; Table 4). Of par-
r interest, the majority of amplifications at 8p23.1
, MYC and 4 other genes), 8p23.1 (31.7%, CTSB5 others), 7q21 (30.8%, CDK6 and 4 others),were focal amplifications encompassing the CTSB gene(Figs. 1B and 3C).
BM to EAC. X axis denotes samples, Y axis denotes chromosomed black, centromere position.
Cancer Prev Res; 3(9) September 2010 1181
. ebruary 1, 2020. © 2010 American Association
Discu
Malationdysplaand thdistribaberrafor a ntissueity, diinteraries ofappliehighealteratour hi3p14.LGD sfrequecontraand msurrourencesuppoinstabplasiaEAC stionsa quagenomCFS
hibitreplicidentgenesjorityIMMP30, 31new C4q22OtherCFS gMGC4We coEACinvolvonlyknowEAC iicallysure dshownstress,Ch
(10–1to bethat m
Fig. 3.A, 3p14amplificPanel 1absolutnormalregionstissuescopiesvalues of 0. A negative log2R value indicates chromosome loss and apositive value suggests chromosome amplification.
Gu et al.
Cancer Prev Res; 3(9) September 20101182
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
ssion
ny previous studies have shown an increasing accumu-of chromosomal aberrations during the metaplasia-sia-carcinoma sequence of EAC development (9–14)e pathogenesis of other malignancies (25–27). Theution and frequency spectrum of each chromosomaltion in the literature has been inconsistent (17, 20)umber of reasons, including assay technology, normalcontamination, small sample sizes, tumor heterogene-fferent environmental exposures, gene/environmentctions, and population stratifications. Using a large se-premalignant andmalignant tissues, the current studyd high-density SNP array technology to produce thest-resolution genome-wide catalogue of chromosomalions in BMthroughEAC todate. The clear advantage ofgh-resolution arrays was shown when the losses of the2 and 16q23.1 regions in samples from the BM andtages were observed exclusively within the two mostntly activated CFS regions in the human genome. Inst, 9p deletions in both dysplastic (LGD and HGD)alignant (EAC) biopsies often involved large regionsnding the p15 and p16 loci. The predominant occur-of these three chromosomal losses in the LGD stagerts the notion that p15 and/or p16 genes and geneticility are early driving forces in the development of dys-. With further malignant progression to the HGD andtages, the spectrum and size of chromosomal aberra-gradually increased. By the EAC stage, approximatelyrter (17% losses and 8% amplifications) of the wholee exhibited aberrations.regions are large, unstable genomic regions that ex-
increased chromosomal gaps and breaks when DNAation is partially inhibited (28–31). The single genesified from this study (Table 3) were mostly large(from ∼500 kb to over 2 Mb) that included the ma-of previously confirmed CFS genes (FHIT, PARK2,2L, CNTNAP2, CTNNA3, WWOX, and DMD; refs.). In addition, our data indicated several potentialFS genes, including KIF26B at 1q44, MGC48628 at.1, GMDS at 6p25.3, and C20orf133 at 20p12.1.s have previously hypothesized MGC48628 to be aene (28). Homozygous deletion of ∼500 kb within8628 was recently reported to be found in EAC (20).nfirmed the same deletion in both dysplastic andtissues. Whether these CFS genes are specificallyed in the development of a subset of EACs or are“hitchhikers” (32) remains to be studied. To ourledge, the occurrence of CFS regions in BM ands more extensive than in other cancers, which biolog-may be attributable to the constant bile acid expo-uring gastrointestinal reflux. Previous studies havethat bile acid induces inflammation and oxidativewhich in turn activates CFS regions (33, 34).
romosomal loss at 5q occurs frequently in EAC6). The APC gene at 5q21 has been hypothesized
the candidate TSG in EAC (35); however, we foundRepresentative pictures of chromosome loss and amplification..2 loss; B, 9p LOH and 9p21 homozygous deletion; C, 8p23.1ation. Panel 4 shows the allele frequencies of normal DNA.shows the allele frequencies of disease tissue. Panel 3 shows thee allele frequency differences between disease (panel 1) and(panel 4) tissues (|dAllelefreq|). Regions of |dAllelefreq| > 0 indicateof aberrations. Panel 2 shows the log2R ratio of disease to normal(measuring changes in intensity from what is expected for twoof each locus). The regions without aberrations should have log2R
ost of the 5q losses were around the PDE4D gene
Cancer Prevention Research
. ebruary 1, 2020. © 2010 American Association
(encocludinand fiilar d
(36)in lunanismquencadenoThe
21q22three(Tabletranscsubunformsand pRUNX(37, 3esophreportRUNXtissuemay hincludWe
stages17p13frequewith pof methat paneupThe
tion rcontation gEGFRampli8p23.exprescodesmiceeratioexhiboverexvious(43).We
rationconsisthereaberra(Fig. 1tionsampleaberramososomeaberra
Fig. 4.amplificchromotrend =chromotrend =aberratand P f
SNP Array of Barrett's Esophagus and EAC
www.a
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
ding a cyclic AMP phosphodiesterase) at 5q11.2, in-g focal losses within PDE4D in two LGD samples
ions per tissue in each stage (P for trend = 2.64 × 10−6 for lossor trend = 1.26 × 10−8 for amplification).
ve EAC samples. Nancarrow et al. (20) reported sim-eletions in EAC samples. Interestingly, Weir et al.
theseable a
acrjournals.org
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
found homozygous focal deletions within PDE4Dg adenocarcinoma. The underlying biological mech-of this homozygous deletion, its functional conse-e, and its pathologic impact on esophageal and lungcarcinoma warrant further investigation.second potential candidate TSG in EAC is RUNX1 at.12. Focal loss of RUNX1 occurred in two LGD andHGD samples, and 21q loss was frequent in EAC2). RUNX1 belongs to the Runt domain family ofription factors consisting of three DNA binding αits (RUNX1, RUNX2, and RUNX3), each of whichheterodimers with the common β subunit CBFβlays pivotal roles in neoplastic progression (37).1 (AML1) is a well-established TSG in leukemia8), whereas RUNX3 is an established TSG in gastric,ageal, and other solid tumors (37, 39). A previousshowed a dramatic downregulation of RUNX1 and3 in gastric tumors compared with adjacent normals (40). Taken together, these data suggest that RUNX1ave tumor suppressor functions in solid tumors,ing gastric and esophageal cancers.found loss of the 9p21 region to occur in the early(BM and LGD), whereas loss of the p53 locus atoccurred in the later stages and was one of the mostnt events in the EAC stage. These data are consistentrevious reports that p16 drives the early progressiontaplasia and that p53 inactivation is a late eventermits further genomic instability and promotesloidy (41).sizes of the small overlapping regions of amplifica-anged from <500 kb to >10 Mb. These regionsined some well-characterized oncogenes or prolifera-enes, such as Myc, CDK6, ZNF217, KRAS, CCND1,, PIK3CA, and VEGF. The most frequent regions offication in EAC were 8q24 (containing Myc) and1 (containing CTSB; Table 4). Myc is the most over-sed oncogene in human cancers, whereas CTSB en-a lysosomal cysteine proteinase. CTSB-deficient
have been shown to have reduced tumor cell prolif-n, and cancer cells lacking CTSB have been shown toit resistance to apoptosis (42). Furthermore, thepression of CTSB mRNA and protein has been pre-ly observed in EAC compared with normal tissuesTherefore, CTSB is a candidate oncogene for EAC.found a gradual increase in chromosomal aber-s during themetaplasia-dysplasia-carcinoma sequence,tent with reports in the literature (9–16). However,were large intersample variations in chromosomaltions among tissues with the same histologic stage), with some samples having no or very few aberra-and some with large numbers of aberrations. For ex-, among 42 EAC samples, 3 samples did not have anytions and 1 sample had only one aberration on chro-me 8q (Fig. 1). Akagi et al. (15) recently reported thatof the EAC samples did not have any chromosometions and suggested that the low tumor content in
Increased level of chromosomal aberrations (losses andations) with increasing histologic stage. A, mean numbers ofsome arm aberrations (P for trend = 4.82 × 10−7 for loss and P for1.35 × 10−9 for amplification). B, mean total sizes (Mb) ofsome aberrations (P for trend = 8.13 × 10−8 for loss and P for1.23 × 10−8 for amplification). C, mean percentage of SNPs with
samples was a possible reason for the lack of detect-berrations. In our study, most EAC samples have
Cancer Prev Res; 3(9) September 2010 1183
. ebruary 1, 2020. © 2010 American Association
abundWe besomegeneittent indetectThe
sectio
separapothemarkewithpremawere
Table t sm ping f ch e lo
Chrom Sta En ize (k G ene ragile
1q44 4352 4376 236. K 548, FRA3p14.2 60383 60475 92. ,502, FRA4q22.1 91643 91880 237. MG 546, FRA5q11.2 58523 58711 187. P ,019,6p25.3 1206 1658 451. G 621, FRA6q26 6195 6224 285. P .380, FRA7q31.1 1064 1101 372. IM 899, FRA7q35 4530 4582 517. CN ,304, FRA9p21.3 21987 22122 134. C 6,10q21 68040 68345 305. C ,776, FRA116q23 77146 77164 18. W ,113, FRA117p13 7413 7736 322. 53 an 198 (p20p21qXp2
NOT
Tab
Chr
3q25p16p26p27p27p17q28p28q211q12p12q13q17q18q20q
NOT
Gu et al.
Cance1184
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
ant tumor cells (>50% tumor cells in the specimen).lieve that the lack of chromosomal aberrations inEAC samples are likely due to inherent tumor hetero-y; however, we cannot rule out that low tumor con-occasional tumor samples may have resulted in no
able aberrations.
E: Only includes regions of ∼500 kb and smaller.
re are a few limitations to this study. This is a cross-nal study in which all the disease samples were from
sectioand g
le smal ing re mplifi
om Sta En Size (k
6 17323 18658 13,353.2 36348 44184 7,831.3 30559 31040 481.1 41746 44753 3,002.2 7950 2280 1,481.2 54471 56699 2,221.2 92102 92744 643.1 11194 11897 704.2 12766 12922 1,5513 69152 69957 8012 25029 25546 5114 65978 67060 1,0814 53356 65178 11,82
E: Only include regions with >15% frequency in EAC.
r Prev Res; 3(9) September 2010
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
te individuals; therefore, we could not test the hy-sis that chromosomal aberrations may be useful bio-rs for predicting the development of EAC in patientsdysplasia. Another related limitation is that all thelignant tissues (BM, LGD, and HGD) in this studyobtained from patients who had EAC. Other cross-
3. Recurren
all overlap regions o romosomnal biene e
. ebru
ss
omarker studies (e.g., uxpression array) have
Freq (%)
HGD EAC
15.0 20.50.0 20.00.0 16.70.0 16.70.0 25.60.0 25.60.0 30.8
15.0 31.75.0 36.60.0 16.75.0 19.55.0 19.5
10.0 22.5
Cance
ary 1, 2020. © 2010 Am
sing SNP array abeen conducted
PIK3C
VEGF
EGFRCDKCTSBMYC
CCNDKRASMDM
r Prevention Re
erican Associat
osome
rt d S b) enes G size F site2
9398 2 5708 3 IF26B 142 1I 444 589 1 FHIT 1 089 3B 255 768 5 C48628 974 4F 521 295 8 DE4D 1 680 805 389 6 MDS 806 6B1
8628 1 3987 4 ARK2 1 352 6E 1 5504 1 7480 0 MP2L 238 7G 1 2806 1 0585 8 TNAP2 2 634 7I872
076 2 DKN2B 411 .3 186 373 2 TNNA3 1 019 0D .1 033 753 7 WOX 1 014 6D .1 608 254 6 p d 22 others 19, 53) .1 654 049 4 0orf133 2 697 0B 12 14825 15018 192. C2 ,057, FRA222.12 35077882 35179895 102.0 RUNX1 261,5441.1 31612560 31712320 99.8 DMD 2,214,919 FRAXC
nalysisin the
4. Recurrent
l overlapp gions of a cationosome
rt d b) uency Genes0050
0274 0 A and 93 genes -p12 269 542 6 53 genes 3 883 288 0 24 genes737
714 7 and 69 genes 19 133 5 25 genes 518 692 8 and 33 others-q21.3
346 573 2 6 and 4 others 870 260 2 and 15 others1
0810 0171 9 and 4 others .2-q13.3 144 414 5 1 and 10 others .1 283 789 8 and 6 others .3 417 279 2 1 and 8 others .3-q21.32 272 858 3 30 genes .2 007 822 6 38 genes 21 36045 36600 55 10.0 19.511.2 17931784 18382880 451 15.0 17.5 CTAGE1 and 3 others13.2 50951694 51913648 962 0.0 28.6 ZNF217 and 3 others
search
ion
esophinvolvIt canto idechanglignanmolecproaclignanversusstudiepatiensuch sAn
eventsries, lan esteach stion tdeletinormthe inwas dlower9p, anin thiimpoto det
opmeeach sIn c
genomdifferEAC.instabrencedriveand thtionaEAC.bilityassoci
Discl
No p
Grant
Natia MultiM.D. Athe Duat M.DCantu F
Refe1. Ga
202. Jem
2003. Ho
vieAvNo
4. Lagsiz
5. En224
6. ReoeCa
7. Sh85
8. ScmaGa
9. WuHacaAm
10. Rielign316
11. Waancar
12. Li X
SNP Array of Barrett's Esophagus and EAC
www.a
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
agus and other organs (15, 44, 45); like ours, someed assessments of various tissues of cancer patients.be argued that this approach increases the potentialntify high-risk markers because it detects moleculares (including frequency of a change) in prema-cy that accompanies cancer (versus premalignantular changes in noncancer patients). Thus, the ap-h can complement cross-sectional studies of prema-cy biomarkers (in premalignancy-only patients)cancer biomarkers and can complement prospectives of cancer risk in Barrett's or other premalignancyts by identifying candidate high-risk markers fortudies (46).additional limitation is that we grouped geneticthat may be biologically distinct into two catego-
osses and amplifications, because we did not haveimate of the fraction of normal tissues present inample and it is difficult to separate different aberra-ypes (e.g., copy neutral LOH, LOH, or homozygouson) based on log2R ratio when there is substantialal cell contamination. However, it is not likely thatcreasing trend of overall chromosomal aberrationsue to higher percentages of normal tissues in thehistologic stages because the specific losses of 3p,d 16q were actually higher in LGD than EAC tissuess study. Nevertheless, biologically, it would be
rtant to differentiate different types of aberrationsermine the molecular mechanism behind the devel-Recepublish
ces in Barrett's adenocarcinoma and the metaplasia-dysplasia-cinoma sequence. Am J Pathol 2000;156:555–66., Galipeau PC, Sanchez CA, et al. Single nucleotide polymorphism-
baityCa
13. Lapretio
14. Panuca
15. Akdiseso
16. Paad
17. KoTh20
18. Fitoe
19. Peproge
20. Nanusin
21. MothePa
22. Sinceon
acrjournals.org
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
nt of EAC. Future studies are needed to distinguishpecific chromosomal aberration.onclusion, the present study provided a high-resolutione-wide catalogue of chromosomal aberrations at
ent stages of histologic progression from BM toThis study also provided strong evidence that geneticility, as evidenced by the early and extensive occur-of microdeletions in CFS regions, and 9p21 lossthe early progression from metaplasia to dysplasiaat p53 is critical in carcinoma development. Addi-lly, we identified RUNX1 as a potential TSG forFinally, we showed that overall chromosomal insta-index and specific chromosome aberrations may beated with neoplastic progression in BM patients.
osure of Potential Conflicts of Interest
otential conflicts of interest were disclosed.
Support
onal Cancer Institute grants CA111922, CA127672, and CA129906;disciplinary Research Program grant from The University of Texasnderson Cancer Center; Premalignant Genome Atlas Program ofncan Family Institute for Cancer Prevention and Risk Assessment. Anderson Cancer Center; and Caporella, Park, Smith, Dallas, andamilies and Rivercreet Foundation.
ived 12/16/2009; revised 05/21/2010; accepted 05/24/2010;ed OnlineFirst 07/22/2010.
rencesrcia M, Jemal A, Ward EM, et al. Global cancer facts & figures07. Atlanta (GA): American Cancer Society; 2007.al A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics,9. CA Cancer J Clin 2009;59:225–49.rner MJ, Ries LAG, Krapcho M, et al. SEER cancer statistics re-w, 1975-2006. Bethesda (MD): National Cancer Institute; 2009.ailable from: http://seer.cancer.gov/csr/1975_2006/, based onvember 2008 SEER data submission, posted to the SEER web site.ergren J. Adenocarcinoma of oesophagus: what exactly is thee of the problem and who is at risk? Gut 2005;54 Suppl 1:i1–5.zinger PC, Mayer RJ. Esophageal cancer. N Engl J Med 2003;349:1–52.id BJ, Li X, Galipeau PC, Vaughan TL. Barrett's oesophagus andsophageal adenocarcinoma: time for a new synthesis. Nat Revncer 2010;10:87–101.aheen NJ, Richter JE. Barrett's oesophagus. Lancet 2009;373:0–61.hnell TG, Sontag SJ, Chejfec G, et al. Long-term nonsurgicalnagement of Barrett's esophagus with high-grade dysplasia.stroenterology 2001;120:1607–19.TT, Watanabe T, Heitmiller R, Zahurak M, Forastiere AA,
milton SR. Genetic alterations in Barrett esophagus and adeno-rcinomas of the esophagus and esophagogastric junction region.J Pathol 1998;153:287–94.gman PH, Vissers KJ, Alers JC, et al. Genomic alterations in ma-ant transformation of Barrett's esophagus. Cancer Res 2001;61:4–70.lch AK, Zitzelsberger HF, Bruch J, et al. Chromosomal imbal-
sed genome-wide chromosome copy change, loss of heterozygos-, and aneuploidy in Barrett's esophagus neoplastic progression.ncer Prev Res 2008;1:413–23.i LA, Paulson TG, Li X, et al. Increasing genomic instability duringmalignant neoplastic progression revealed through high resolu-n array-CGH. Genes Chromosomes Cancer 2007;46:532–42.ulson TG, Maley CC, Li X, et al. Chromosomal instability and copymber alterations in Barrett's esophagus and esophageal adeno-rcinoma. Clin Cancer Res 2009;15:3305–14.agi T, Ito T, Kato M, et al. Chromosomal abnormalities and novelease-related regions in progression from Barrett's esophagus tophageal adenocarcinoma. Int J Cancer 2009;125:2349–59.ulson TG, Reid BJ. Focus on Barrett's esophagus and esophagealenocarcinoma. Cancer Cell 2004;6:11–6.ppert LB, Wijnhoven BP, van Dekken H, Tilanus HW, Dinjens WN.e molecular biology of esophageal adenocarcinoma. J Surg Oncol05;92:169–90.zgerald RC. Molecular basis of Barrett's oesophagus andsophageal adenocarcinoma. Gut 2006;55:1810–20.iffer DA, Le JM, Steemers FJ, et al. High-resolution genomicfiling of chromosomal aberrations using Infinium whole-genomenotyping. Genome Res 2006;16:1136–48.ncarrow DJ, Handoko HY, Smithers BM, et al. Genome-wide copymber analysis in esophageal adenocarcinoma using high-densitygle-nucleotide polymorphismarrays. Cancer Res 2008;68:4163–72.ntgomery E, Bronner MP, Goldblum JR, et al. Reproducibility ofdiagnosis of dysplasia in Barrett esophagus: a reaffirmation. Humthol 2001;32:368–78.
gletary SE, Greene FL, Sobin LH. Classification of isolated tumorlls: clarification of the 6th edition of the American Joint CommitteeCancer Staging Manual. Cancer 2003;98:2740–1.Cancer Prev Res; 3(9) September 2010 1185
. ebruary 1, 2020. © 2010 American Association
23. OlsmeBio
24. IluthInfHU
25. Racancer
26. Weca
27. BeJ P
28. Du200
29. LukAc
30. Mcranca
31. Smlarg48–
32. MaSep164
33. Zhugen
34. CohypnoMo
35. Huinvca
36. Wead
37. ItogeCa
38. Pecia
39. Lilos11
40. Satiothe22
41. Baline
42. Vaatiin
43. Huresca
44. WaBaad
45. TsexCa
Gu et al.
Cance1186
Do
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265
hen AB, Venkatraman ES, Lucito R, Wigler M. Circular binary seg-ntation for the analysis of array-based DNA copy number data.statistics 2004;5:557–72.mina Technical Bulletin. Whole-genome genotyping withe Sentrix® HumanHap300 Genotyping BeadChip and theinium™ II Assay. Available from: www.servicexs.nl/blobs/illumina/MANHAP300TechBull.pdf.jagopalan H, Nowak MA, Vogelstein B, Lengauer C. The signifi-ce of unstable chromosomes in colorectal cancer. Nat Rev Can-2003;3:695–701.ir B, Zhao X, Meyerson M. Somatic alterations in the humanncer genome. Cancer Cell 2004;6:433–8.ll DW. Our changing view of the genomic landscape of cancer.athol 2010;220:231–43.rkin SG, Glover TW. Chromosome fragile sites. Annu Rev Genet7;41:169–92.usa T, Fryns JP. Human chromosome fragility. Biochim Biophysta 2008;1779:3–16.Avoy S, Ganapathiraju SC, Ducharme-Smith AL, et al. Non-dom inactivation of large common fragile site genes in differentncers. Cytogenet Genome Res 2007;118:260–9.ith DI, Zhu Y, McAvoy S, Kuhn R. Common fragile sites, extremelye genes, neural development and cancer. Cancer Lett 2006;232:57.ley CC, Galipeau PC, Li X, Sanchez CA, Paulson TG, Reid BJ.lectively advantageous mutations and hitchhikers in neoplasms:6 lesions are selected in Barrett's esophagus. Cancer Res 2004;:3414–27.Y, McAvoy S, Kuhn R, Smith DI. RORA, a large common fragile sitee, is involved in cellular stress response. Oncogene 2006;25:2901–8.quelle A, Toledo F, Stern S, Bieth A, Debatisse M. A new role for
oxia in tumor progression: induction of fragile site triggering ge-mic rearrangements and formation of complex DMs and HSRs.l Cell 1998;2:259–65.46. Madic38
r Prev Res; 3(9) September 2010
for Cancer Researchon Fcancerpreventionresearch.aacrjournals.org wnloaded from
ang Y, Boynton RF, Blount PL, et al. Loss of heterozygosityolves multiple tumor suppressor genes in human esophagealncers. Cancer Res 1992;52:6525–30.ir BA, Woo MS, Getz G. Characterizing the cancer genome in lungenocarcinoma. Nature 2007;450:893–8.Y. RUNX genes in development and cancer: regulation of viralne expression and the discovery of RUNX family genes. Advncer Res 2008;99:33–76.rry C, Eldor A, Soreq H. Runx1/AML1 in leukemia: disrupted asso-tion with diverse protein partners. Leuk Res 2002;26:221–8.QL, Ito K, Sakakura C, et al. Causal relationship between thes of RUNX3 expression and gastric cancer. Cell 2002;109:3–24.kakura C, Hagiwara A, Miyagawa K, et al. Frequent downregula-n of the runt domain transcription factors RUNX1, RUNX3 andir cofactor CBFB in gastric cancer. Int J Cancer 2005;113:1–8.rrett MT, Sanchez CA, Prevo LJ, et al. Evolution of neoplastic cellages in Barrett oesophagus. Nat Genet 1999;22:106–9.siljeva O, Korovin M, Gajda M, et al. Reduced tumour cell prolifer-on and delayed development of high-grade mammary carcinomascathepsin B-deficient mice. Oncogene 2008;27:4191–9.ghes SJ, Glover TW, Zhu XX, et al. A novel amplicon at 8p22-23ults in overexpression of cathepsin B in esophageal adeno-rcinoma. Proc Natl Acad Sci U S A 1998;95:12410–5.ng S, Zhan M, Yin J, et al. Transcriptional profiling suggests thatrrett's metaplasia is an early intermediate stage in esophagealenocarcinogenesis. Oncogene 2006;25:3346–56.afrir D, Bacolod M, Selvanayagam Z, et al. Relationship of genepression and chromosomal abnormalities in colorectal cancer.ncer Res 2006;66:2129–37.ley CC, Galipeau PC, Finley JC, et al. Genetic clonal diversity pre-
ts progression to esophageal adenocarcinoma. Nat Genet 2006;:468–73.Cancer Prevention Research
. ebruary 1, 2020. © 2010 American Association
2010;3:1176-1186. Published OnlineFirst July 22, 2010.Cancer Prev Res Jian Gu, Jaffer A. Ajani, Ernest T. Hawk, et al. High-Density Single Nucleotide Polymorphism Array AnalysisBarrett's Esophagus and Esophageal Adenocarcinoma: A Genome-Wide Catalogue of Chromosomal Aberrations in
Updated version
10.1158/1940-6207.CAPR-09-0265doi:
Access the most recent version of this article at:
Material
Supplementary
0265.DC1
http://cancerpreventionresearch.aacrjournals.org/content/suppl/2010/07/26/1940-6207.CAPR-09-Access the most recent supplemental material at:
Cited articles
http://cancerpreventionresearch.aacrjournals.org/content/3/9/1176.full#ref-list-1
This article cites 43 articles, 11 of which you can access for free at:
Citing articles
http://cancerpreventionresearch.aacrjournals.org/content/3/9/1176.full#related-urls
This article has been cited by 15 HighWire-hosted articles. Access the articles at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
.http://cancerpreventionresearch.aacrjournals.org/content/3/9/1176To request permission to re-use all or part of this article, use this link
for Cancer Research. on February 1, 2020. © 2010 American Associationcancerpreventionresearch.aacrjournals.org Downloaded from
Published OnlineFirst July 22, 2010; DOI: 10.1158/1940-6207.CAPR-09-0265